IRC log of #maemo for Sunday, 2012-07-01

Sicelo:/00:00
Sicelomaybe i should try a reboot00:03
*** ArGGu^^ has joined #maemo00:03
Sicelomerlin1991: what's the actually mfe app .. i want to try via command line if possible00:07
*** Mrick0363 has quit IRC00:09
*** atlas has joined #maemo00:11
*** croppa has joined #maemo00:16
*** Dr_I has joined #maemo00:18
*** ced117_ has quit IRC00:19
*** nmjnb has quit IRC00:22
merlin1991Sicelo: I typed all my stuff into the ui and then it semi works00:24
*** eijk has quit IRC00:24
Dr_Ifine Vr ofency on nokist time00:25
Dr_Igood working-class on vr00:26
Sicelothe thing is .. the UI doesn't even attempt to send any packets. weird00:27
DocScrutinizerDr_I: is that just me or is every single post of yours like a random selection of words and TLAs from dictionary?00:28
*** _federico3 has joined #maemo00:28
DocScrutinizerDr_I: I'm already about to conclude you're a poorly written script00:29
Siceloi expect to get some connection attempt (even if i possibly have a non-supported server). in this case, nothing at all is sent. it's like the UI isn't even aware of the existence of a network connection00:30
Sicelothen again, the 'activesync .. pidfile was not' created makes me wonder if the actual binary is even able to run :/00:30
DocScrutinizeranybody ever chatted with Dr_I ? a single meaningful word?00:30
DocScrutinizerDr_I: last chance to utter some sensible thing to pass turing test, otherwise I'll +q you00:31
*** trbs has quit IRC00:33
*** ChanServ sets mode: +o DocScrutinizer00:33
*** DocScrutinizer sets mode: +q Dr_I!*@*00:33
DocScrutinizerDr_I: ping me per /query if you decide to act like a human00:34
*** ChanServ sets mode: -o DocScrutinizer00:34
merlin1991Sicelo: afaic the active sync thingy is what mfe runs00:35
Sicelohttp://paste.debian.net/177136/ .. that's the log00:35
merlin1991Sicelo: also when I hit my email and pw settings and press the button to continue (don't remember exaclty, has been some time) it did complain a lot about bs like no valid chain for the server certificate, invalid login (typos) ...00:35
merlin1991so there had to be some connectivity00:36
*** Vlad_on_the_road has quit IRC00:36
Siceloi can see from tmo that the certificate stuff is known to have issues, but i won't even get that far, as MfE doesn't even attempt to contact any server00:36
merlin1991Sicelo: I don't have my n900(s) here atm, but when I'm back in vienna (sunday night) I can investigate what should happen if you add the stuf00:37
DocScrutinizerSicelo: do you have connectivity up during that process?00:37
*** flo_lap has joined #maemo00:38
Siceloyes DocScrutinizer.00:38
DocScrutinizerhmm, then I dunno00:38
Sicelowireshark just remains silent the whole time00:38
DocScrutinizertried strace?00:38
Sicelonop. let me00:38
Sicelobut i might need some hand-holding there .. strace isn't a tool i'm so familiar with00:39
DocScrutinizerdon't forget the -f ork option00:39
DocScrutinizerin the simplest form strace can either start a process (strace date)00:40
DocScrutinizeror attach to a running process (strace -p `pidof less`)00:40
DocScrutinizer-f (iirc) means "also follow child processes that might get forked"00:40
DocScrutinizerif nothing is given for a filter, szrace will simply trace *ALL*00:41
DocScrutinizeryou might want to redirect the diagnostic output to a file, so you could analyze later (-o iirc)00:42
DocScrutinizerthere's a nice terse howto on wiki.maemo.org00:43
DocScrutinizerhttp://wiki.maemo.org/Documentation/devtools/maemo5/strace00:43
merlin1991oh god -o00:47
merlin1991tried to use -o with wget one day wondering wtf went wrong00:47
merlin1991untill I realized, wget has -o and -O00:48
DocScrutinizerhmm00:48
merlin1991-o redirects stdout, -O redirects file output00:48
DocScrutinizerI'm all but sure about -o of strace, pretty please look it up in strace --help or man strace00:48
merlin1991(for wget that is)00:49
*** messerting has joined #maemo00:49
DocScrutinizerSicelo: one catch about strace is: all file operations (you know *unix* almost all is a file) are against filehandles which are numeric. You have to look up the open() call that associates the filehandle to a real useful path and filename00:51
Sicelochmod 666 /var/lock/as-daemon.pid seems to be getting me somewhere :P00:51
Siceloindeed, now i'm on next screen ;)00:51
*** sq-one has quit IRC00:51
DocScrutinizerstale lock? just delete it00:52
*** PeterWolf has joined #maemo00:53
DocScrutinizerbtw it might be interesting to do `cat  /var/lock/as-daemon.pid`00:55
DocScrutinizerand ps | grep `cat  /var/lock/as-daemon.pid`00:55
DocScrutinizerand wtf is as-daemon anyway? :-D00:56
Siceloactive sync daemon :P00:57
DocScrutinizeraaaah :-D00:57
RiDalso known as A W E O00:58
DocScrutinizerso a potential explanation is simple: as-daemon stalled for some reason, and left a stale lockfile  /var/lock/as-daemon.pid. This made any subsequent startup of same daemon abort and fail00:58
DocScrutinizert900:~# ps | grep `cat /var/lock/as-daemon.pid`01:01
DocScrutinizer20804 user     24392 S    /usr/sbin/as-daemon -D --pidfile=/var/lock/as-daemon.01:01
*** qwazix_N950 has quit IRC01:05
*** _federico3 has quit IRC01:17
*** retro|cz has joined #maemo01:20
*** flo_lap is now known as florian01:27
*** muelli has joined #maemo01:27
*** NIN101 has quit IRC01:29
freemangordonzeq: did you try it?01:31
zeqfreemangordon: yes, it worked well :)01:37
zeqwe'll have to let romaxa know01:37
freemangordongood. zeq, I found a problem with your zoom functionality, it does not reflow the text01:37
freemangordonI told him01:37
zeqah good!01:38
freemangordonhe was not impressed :D01:38
zeqlol01:38
zeqwhy?01:38
zeqnot the correct solution?01:38
freemangordon[19:34] <freemangordon> romaxa: orientation change support for maemo5 is fixed now, most probably either me or zeq will file a bug (with patch). On the other hand I tried GLES support, it works, but spits some errors (couldn't create EGL image: ERROR (0x300c)) and is slower than without it. I suppose GLES support is still not finished.01:38
freemangordon[20:34] <romaxa> freemangordonÐ: hmmm possibly.. EGLImage might not work, but there are definitely works simple TFP bindImage to X surface... also thebesLayers tiled version need to be enabled when it is ready for IPC version, that should make it works just great01:39
freemangordon[20:34] <romaxa> freemangordon: also in order to remove that error you can try to set in GLContextProviderEGL.cpp gUseBackingSurface = false01:39
nox-http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-today01:39
freemangordonzeq: BTW we are NOT using HW acceleration01:39
zeqgood potential for more speedups then01:40
zeqas for the zoom, it's a bug/feature in the mozilla code01:40
freemangordonwell, not unless romaxa (or some other GL guru) helps01:40
zeqI just enabled the buttons01:40
freemangordonzeq: but it works on double-click01:40
freemangordon(ther is an option for that)01:41
zeqI guess it's different functionality.  I can't remember what the old gtk version did01:41
freemangordonwhatever, without text reflow, zoom is unusable in portrait01:41
freemangordonAIUI01:42
zeqI did look through the code for where the keybindings are for zoom, but couldn't find them01:42
zeqI don't know if there is some other zoom mode that could be hooked up01:43
*** PeterWolf has quit IRC01:43
DocScrutinizernox-: what's the bottom line?01:43
freemangordoni am grepping for OnMouseDoubleClickEvent through the whole code right now, there should be some function called that reflows01:43
nox-DocScrutinizer, ntp vs leap second vs kernel bug apparently01:44
zeqfreemangordon: that's what I was thinking01:44
nox-ntpd01:44
DocScrutinizerhaha, leap second01:44
DocScrutinizergreat!01:44
nox-indeed01:44
* nox- glad he uses freebsd for this01:44
*** erstazi has quit IRC01:44
zeqfreemangordon: I'm going to have to head to bed soon...01:45
DocScrutinizerI could easily see ntpd et al crashing on 02:00:6001:45
freemangordonzeq: np, i will look at it01:45
zeqdid romaxa's comments about gles make any sense to you at all?01:45
nox-DocScrutinizer, yeah but it seemed to take the kernel with it01:45
DocScrutinizeralso the leap-second announce flag in services like DCF77 is a rarely used and thus untested feature01:46
freemangordonzeq: partially. i tried his suggestions, no difference. and with enabled hw acceleration fennec crashes every now and then01:46
zeqegl acceleration or render..?01:47
freemangordonyep01:47
freemangordonrender01:47
zeqno faster?01:47
nox-quote: This is caused by a livelock when ntpd calls adjtimex(2) to tell the kernel to insert a leap second.01:47
freemangordonwell, it peaks at about 38 fps, but stutters01:47
zeqfreemangordon: what do we get without?01:48
freemangordon(there is an option in about:config to enable HW acceleration and fps print)01:48
zeqok01:48
freemangordonzeq: rephrase please01:48
zeqfps without hw accel01:48
freemangordonno idea, it works only when HA accel is enabled01:49
freemangordon*HW01:49
zeqthe instability could be coming from anywhere01:49
freemangordon[21:06] <freemangordon> romaxa: actually the fps goes up to 34-36 from time to time, but the whole thing stutters01:49
freemangordon[21:07] <freemangordon> romaxa: thanks01:49
freemangordon[21:07] <romaxa> freemangordon: yep it is becasue slow GL upload01:49
*** dhbiker has quit IRC01:49
freemangordonwhatever it means01:50
zeqit can't be uploading on each character?01:50
freemangordonnnah, i suspect that means that texture upload is slow01:51
freemangordonfrom memory to GPU01:51
zeqhe could be right.  I've no idea how good it is.01:51
freemangordonNFC actually, GL is a mistery for me for most of its part01:52
freemangordonwell, maybe if we pester romaxa often enough he will finally take a look at it :)01:52
zeqyou never know.. he did say he *might* set up a maemo sdk and give it a try01:53
freemangordonzeq: but better you do it, I am famous with my diplomatic skills :D01:53
zeq:)01:53
zeqfreemangordon: I'm off to bed now, catch you tomorrow?01:54
freemangordonanyway, I think the first set of patches should be ready once I fix that text reflow issue01:54
freemangordonsure01:54
zeqA working set of patches has got to help our chances of getting some help :)01:55
freemangordonyep01:55
freemangordonseems there are lots of GLES gurus on mizzila bugtracker01:56
freemangordon*mozilla01:56
*** marainein has quit IRC01:57
*** markinfo has quit IRC01:59
*** Hurrian has joined #maemo02:01
*** zeq has quit IRC02:04
DocScrutinizernox-: thanks02:04
nox-np02:05
ShadowJKAnyone want to buy a Sandisk microsdhc 32gigabyte class 4? :-)02:12
*** GuySoft has quit IRC02:15
SpeedEvilIs this the one approaching 1iops in some tests?02:16
ShadowJK:)02:16
*** Hurrian has quit IRC02:16
*** piggz has quit IRC02:17
ShadowJK4 threads sequential write 4M block size: 8 meg/sec02:17
DocScrutinizeris sandisk not what sandisk used to be, anymore?02:20
ShadowJKseems so02:20
ShadowJKBut also the capacity difference might be significant factor02:21
DocScrutinizeror is this one of the fake sandisk chinese counterfeits02:21
ShadowJKI've benchmarked the old C2 and C4 sandisks at 8G capacity, and this is new C4 at 32G02:21
DocScrutinizermaybe even a 4GB one that just tells "hey I'm 32"02:21
ShadowJKWell, the manfid etc check out, and the controller is new, because it actually reports a "preferred erase size"02:22
ShadowJK(of 4M)02:22
DocScrutinizeryeah, we've seen such things for USB sticks a lot02:22
DocScrutinizersome years ago02:22
DocScrutinizerwouldn't know why they can't do same thing for uSD02:23
*** rzr is now known as rZr02:23
ShadowJKrandom read speed peaks at 18M/sec for large block sizes, and seems mostly limited by command overhead when going towards smaller block sizes02:23
ShadowJKbought from a "real", established webstore, not ebay or dealextreme :P02:24
DocScrutinizerhmm chumby or what's been the name also did02:24
ShadowJKNo they bought Kingston02:25
DocScrutinizeraaah yes, maybe02:25
ShadowJKKingston itself buys random crap and rebrands it02:25
DocScrutinizerwell, on his investigations he found some fake cards anyway, not sure if all were kingston, iirc they were sandisk (who woulld fake kingston anyway?)02:26
*** GeneralAntilles has quit IRC02:26
* DocScrutinizer seems to recall incorrect SD logo02:27
*** GeneralAntilles has joined #maemo02:27
DocScrutinizerwith that gap in the "D"02:27
*** erstazi has joined #maemo02:28
DocScrutinizerhttp://www.bunniestudios.com/blog/?p=91802:28
DocScrutinizermeh, kingston indeed02:28
*** SmilybOrg has joined #maemo02:31
*** SmilybOrg is now known as Guest8702602:32
*** OkropNick has quit IRC02:34
*** SmilyOrg has quit IRC02:35
*** florian has quit IRC02:36
*** GuySoft has joined #maemo02:38
*** ekze has quit IRC02:42
*** FIQ|n900 has joined #maemo02:49
*** valdyn has quit IRC02:52
*** valdyn has joined #maemo02:52
*** Mrick0363 has joined #maemo02:55
*** messerting has quit IRC02:57
*** messerting has joined #maemo02:57
*** udovdh has quit IRC02:57
*** robink has quit IRC02:59
*** uen| has joined #maemo03:02
*** ab has quit IRC03:02
DocScrutinizeranyway pairing a (possibly slightly 'tuned') 32GB controller with a 2 or 4 GB flash seems just too simple to not get done eventually03:04
*** robink_ has joined #maemo03:04
DocScrutinizerbad block management will easily cope with the missing storage03:05
*** uen has quit IRC03:05
*** M4rtinK has quit IRC03:12
*** rZr is now known as rzr03:18
*** Hurrian has joined #maemo03:21
*** messerting_ has joined #maemo03:22
*** messerting has quit IRC03:22
*** messerting_ has quit IRC03:26
*** RiD has quit IRC03:33
*** Mrick0363 has quit IRC03:34
*** brzys has quit IRC03:35
*** githogori has quit IRC03:52
ShadowJKbad block management mostly doesn't exist ime :/03:53
*** rzr is now known as rZr03:54
*** hurbu has joined #maemo03:54
*** ColdFyre has quit IRC03:58
*** atlas has quit IRC04:01
*** ColdFyre has joined #maemo04:01
*** e-yes has quit IRC04:02
*** muelli has quit IRC04:04
*** erkules|away has joined #maemo04:05
DocScrutinizeryou don't think bad pages are marked as bad somewhere, and won't get used anymore?04:06
*** hurbu has quit IRC04:06
DocScrutinizerI'd guess that would drastically complicate and make expensive any working uSD04:07
*** erkules has quit IRC04:07
DocScrutinizeraiui controllers are built in a way they detect and 'discard' and bad pages, and thus by mere over-providing of storage manufacturers make sure their cards will work with the advertised capacity, even while _usually_ some pages on flash chip are defect04:09
DocScrutinizerthe huge difference between NAND and NOR04:10
DocScrutinizers/and bad/any bad/04:10
DocScrutinizermuch the same way modern HDD do the same04:12
*** PeterWolf has joined #maemo04:12
DocScrutinizerthough those still use a direct geometrical (or linear) mapping of blocks, and only remap bad blocks. While a uSD is juggling good pages all around, if only for wear leveling. This wear leveling will probably tag and discard and not use any further any bad page it detects, as anything else seems silly04:14
*** npm_ is now known as npm04:17
DocScrutinizerwell, you're probably right when you take my term "block" literally04:17
DocScrutinizerof course that mapping and tagging will work on nothing smaller than a whole erase page04:18
DocScrutinizerof course I also heard of those cases where card just starts to malfunction as soon as all spare pages got used up04:20
DocScrutinizerbut I'd rather guess that's only happening when you actually try to fill up the whole nominal storage capacity of the card with data, which obviously can't work anymore when the controller runs out of spare pages to use for both wear leveling and substitution of defect pages04:22
DocScrutinizerdepending on smartness of controller algo devels04:24
*** Atarii has quit IRC04:25
*** robink_ is now known as robink04:29
*** robink has joined #maemo04:29
*** nox- has quit IRC04:39
ShadowJKLast person that investigated the low level details concluded there's about 2% reserve pages, and the wear leveling is accidental side effect of something else, and bad block replacement only works within 1000 page chunks04:47
ShadowJKi.e. first 800M gets 40M spare, once that spare is used up the device pretty much dies even if there was otherwise lots spare left elsewhere ond evice04:48
* ShadowJK finds a 4G C10 Transcend..04:50
ShadowJKI wonder why this is not in my spreadsheet04:50
ShadowJKmaybe I started benchmarks and got pissed off at it or osmething..04:50
*** erkules|away has quit IRC05:02
*** r00t|home has quit IRC05:08
*** radic has quit IRC05:09
*** r00t|home has joined #maemo05:10
*** erkules|away has joined #maemo05:15
*** rcg has quit IRC05:18
*** aloril has quit IRC05:19
*** radic has joined #maemo05:22
*** aloril has joined #maemo05:24
DocScrutinizeraah, ok, then bad block management is probably irrelevant for faking higher capacities05:29
*** XDS2010 has joined #maemo05:30
*** rcg has joined #maemo05:33
* kwtm2 weeps. Just verified that 2 out of my four Official Nokia BL-5J batteries (with hologram) are bricked. :P05:44
*** MacDrunk has joined #maemo05:45
*** LaoLang_cool has joined #maemo05:45
HurrianDocScrutinizer, depends on how fake cards do it05:45
Maceri'll be damned05:45
Hurriansome do it via corrupt partition table, others via special controller05:46
Macermy k45 mboard that died a while back has bad caps05:46
*** MacDrunk has left #maemo05:46
Macer4 of them are bulging05:47
*** robink has quit IRC05:49
kwtm2bulging means near end of life?  Similar to rechargeable cellphone batteries?05:50
Macerwell.. the board won't turn on anymore ;)05:51
Macerso i'm guessing that most of them are broken05:51
*** robink_ has joined #maemo05:51
Macerit has like 7 bad caps05:52
Maceroh wait... missed one.. 8 heh05:52
Macerwonder if i can buy some new ones at radio shack05:52
Macerdoes radio shack even sell those types of things anymore?05:52
Macerwhat is odd is that i actually bought 2 of these within like 3 weeks of one another.. one of them works grat05:53
ShadowJKkwtm2, bulging means broken05:53
Macermade in china :-P05:54
Maceri need to find some capacitors and a soldering iron05:54
ShadowJKI've been meaning to solder a broken gigabit 8 port switch05:56
ShadowJKI ordered caps, but hten found other use for them, heh.05:56
ShadowJKThe caps on the board are underspecced according to their label05:57
Macerheh05:58
Macerthe group of caps that are broken are rated all the same05:58
Macerthey are the same type of cap05:59
Macerit's a shuttle k45 board05:59
Macerthere were people winding up with like 8/10 shuttles dying05:59
Macerbeause of the caps05:59
kwtm2ShadowJK: I see.  Same principle as bulging batteries on cell phones, or not related?  Does bulging mean internal explosion?05:59
ShadowJKI think the process is different06:00
ShadowJKmaybe06:00
Macerblah i have to go see about getting the stuff for this06:01
*** dockane_ has joined #maemo06:01
*** dockane has quit IRC06:01
Macerhttp://us.shuttle.com/scgforum/tm.aspx?m=321206:02
Macerthat's the same exact problem mine had ;)06:02
Macerthe 1800uF/6.3v ones are bulging heh06:04
Macercobalt-60 found on petco dog bowls06:06
Macerwtf are these chinese doing?? lol!06:06
*** mase76 has quit IRC06:07
Macerhttp://www.filmcapacitors.com/SeriesMF120606:11
Maceri'm guessing that place won't sell me 6 caps huh? :)06:11
MacerShadowJK: what is a place to get good high quality caps?06:11
Macerbonus points if you can find me caps made in the US :)06:12
ShadowJKI don't think there are any06:14
ShadowJK(any US production of capacitors)06:14
ShadowJKBut like, uh, rubycon and Nichicon?06:16
*** Gh0sty has quit IRC06:19
ShadowJKdigikey mouser or whatever06:19
*** mase76 has joined #maemo06:20
*** Gh0sty has joined #maemo06:21
*** udovdh has joined #maemo06:22
*** atlas has joined #maemo06:23
*** Estel_ has quit IRC06:31
*** robink_ is now known as robink06:38
Macerfriend of mine is going to order me some when he orders parts for work06:39
*** robink has quit IRC06:39
*** robink has joined #maemo06:39
Macerso i can just wait on those... lol.. but yeah.. finding us made caps would be impossible06:39
Macerhe said some place in ohio makes them still06:39
Macerall US caps are like... special order bulk orders only06:39
*** Estel_ has joined #maemo06:40
*** jargon- has quit IRC06:50
*** jargon- has joined #maemo06:57
*** robbiethe1st has joined #maemo07:05
*** valdyn has quit IRC07:13
ShadowJKThis isn't limited to just capacitors07:13
*** luke-jr has quit IRC07:13
ShadowJKJust about anything is so much easier to get from asia than from the us07:14
*** Kilroo has quit IRC07:15
ShadowJKLike, I know some things where a US company designed something to be manufactured in china, and actually has good enough control of the factory in china, the product gets shipped right away to usa, but it's easier and cheaper to have it shipped back to china and sold from china07:16
ShadowJKthan to try do business with the us company07:16
ShadowJKheh07:16
*** dimw1t has joined #maemo07:16
*** atlas has quit IRC07:16
ShadowJKexcept if you live in the us, that is07:17
*** luke-jr has joined #maemo07:19
*** ferdna has joined #maemo07:20
*** valdyn has joined #maemo07:23
*** Kilroo has joined #maemo07:26
*** dhbiker has joined #maemo07:28
*** dimw1t has quit IRC08:00
*** Estel_ has quit IRC08:04
*** Estel_ has joined #maemo08:04
MacerShadowJK: yeah that sucks08:07
Macerthat's most places now08:07
Macerbut every now and then you find some company that is actually in the united states08:08
Macerlol08:08
*** Mrick0363 has joined #maemo08:11
Estel_DocScrutinizer, feel silly continuing this topic, but re IRc and referendum - you're wrong, as usual. While I agree that it would be idiotic idea - and I used it as comparision - actually, chan imperator is working closely with Council, so TMO could - following your twisted logic - make referendum to have Council ask X-Fade about removing You frokm chanop08:12
Estel_and X-fade would do that, per Council request08:12
Estel_again, it's just "mind experiment" don't worry, it was just about showing how You pivot rules to your liking, in my humble opinion, which you have, of course, right to not agree.08:13
Estel_If You think that referendum can be used to remove volunteers form volunteering, it can be used to make anything. It's kinda wrong, as AFAIk, referendum is something started by Council, to see if poeople support some dieas or not.08:14
Estel_after all, council doesn't do anything alone, cause it's not fuckin goverment, despite fact, that almsot everyone is trying to apply political filters.08:14
*** ekze has joined #maemo08:14
Estel_i.e. Council can say "we're doing this and that", but without other volunteers wanting to do that, Council can kiss own ass, if manage to.08:15
Estel_it's all about collaboration.08:15
Estel_due mtot his reason, applying copy& paste from real-world goverments and politics is stupid. Not to mention, that even in real life, You have election every 4-5 years, and goverment can't be forced to go away (in legal ways, not by guns ;) ) even if have only 1% of support08:16
Estel_fortunately, despite how You see it, maemo is also absed on meritocracy. Fact that we have election of *volunteers* doesn't mean it becomes political.08:17
Estel_anyway, whatever, I'm sure that You won't agree, and everyone have better things to do, so... :)08:18
*** ab has joined #maemo08:24
Macerthe woman who plays targaryen in game of thrones is way hotter as a blonde08:35
*** githogori has joined #maemo08:39
*** LaoLang_cool has quit IRC08:42
Estel_Macer, You're talking about it for week already, have You feel into platonical love, or what?08:56
Estel_Better tell me how cooking N900 in oven went08:56
Maceroh... damn. i totally forgot all about that heh08:56
Estel_I knew that. phew.08:56
kwtm2Am about to flash my N900. Remind me again: I only use the -R flag on the second of the two operations (1. flashing firmware and 2. flashing eMMC) whichever is done last, right?  THe page at http://wiki.maemo.org/Updating_the_firmware#Linux is unclear.08:57
kwtm2seen alex08:58
MacerEstel_: i was working on getting some caps for a mboard with bad caps08:58
Macerheh08:58
Macer6 of 8 of them are bulging08:58
Estel_kwtm2, You flash emmc first witohut reboot, then, rootfs08:58
Estel_if You feel masochistc, You may do other way around08:59
Estel_no gain, and every accidentla reboot kills Your integrity08:59
Estel_emmc first without r and disabling it after by removing battery + flashing rootfs with -R later is best way08:59
Estel_Macer, well, caps happen :)08:59
Macerit's an older shuttle k4509:00
Macerthey were notorious for it09:00
Maceri bought another one like 3 weeks after the first one and it is still working like a champ... the first one must have gotten the fake black market caps09:00
Estel_or just bad luck :)09:01
Macermy buddy is getting me some new ones from his job... they are pretty easy to replace09:01
*** npm has quit IRC09:01
Estel_lovely times before everything got made for robots only, eh?09:02
Macerwell... caps are rather large.. somewhat difficult to not make them easy to replace09:02
MacerK45s were awesome tho09:02
Macerwhen the caps didn't explode :-P09:02
Estel_Don't underrate invention of designers :P09:03
Estel_You will see, when we will start to have caps with balls soldering ;)09:04
Estel_just for the sake of it.09:04
*** npm has joined #maemo09:05
*** githogori has quit IRC09:30
kwtm2thx Estel_09:32
kwtm2ok wait...09:33
kwtm2is rootfs firmeware, then?09:33
kwtm2s/firmeware/firmware/09:34
infobotkwtm2 meant: is rootfs firmware, then?09:34
kwtm2is the -R flag for rebooting, then?09:35
*** Pali has joined #maemo09:35
*** ab has quit IRC09:35
kwtm2once I failed to be able to flash firmware, causing problems that have plagued me to this day.  That was the most recent time (although I had flashed it half a dozen times prior to that, some time before).  Hope it works this time.09:39
*** LaoLang_cool has joined #maemo09:41
Estel_kwtm2, yea09:45
Estel_emmC is just partition layout of eMMC09:45
kwtm2Okay, failed again.  Says "missing image".  I wonder about a number of things that are unclear on the wiki.09:45
Estel_rootfs is... rootfs:p09:45
Estel_in fact, nothing is uncleare there, it seems that you're missinterpreting something. Lets try to work through it09:45
kwtm2Estel_: I see that I have 2 files.  (This is from when I flashed a long time ago)09:45
Estel_first of all - if You use flasher without any arguments, it will show you list of options and meanings09:45
kwtm2My two files are called EMMC and FIRMWARE.  So rootfs would be part of the firmware?09:45
Estel_yes.09:46
Estel_RX-51_2009SE_10.2010.13-2.VANILLA_PR_EMMC_MR0_ARM.bin09:46
kwtm2Also, the unclear part: after the flasher exits, is the N900 supposed to sit there unmoving?09:46
Estel_^^^ vanilla eMMc. you don't need it, most of the times.09:46
Estel_wait, wait, first things first09:46
kwtm2ok.  Pls go ahead.09:46
Estel_RX-51_2009SE_20.2010.36-2_PR_COMBINED_MR0_ARM.bin09:46
Estel_^^^ FIASCO image09:46
Estel_containing many trhings for rootfs09:47
Estel_things*09:47
Estel_aka combined09:47
Estel_now, You just do as per wiki - if You want to flash with eMMC image restoration, you feed it by that image...09:47
Estel_flasher-3.5 -F RX-51_2009SE_10.2010.13-2.VANILLA_PR_EMMC_MR0_ARM.bin -f09:47
Estel_first, big -F is about "file"09:48
Estel_it must be path to file09:48
Estel_it depends on system, etc, but it's either relativer or absolute path09:48
Estel_then, -f is "flash"09:48
Estel_or other way around, maybe I've messed upercase and lowercase f - in any case, od as it is in wiki :)09:48
kwtm2The unclear part is some things that I wonder about as a noob, about which I would like reassurance.  Maybe it's just that I need something from my perspective as someone who may possibly brick my n900. :)09:48
Estel_so, you flash eMMC, and then, when it reports that I've finished with success, You remove battery from N90009:49
Estel_I see.09:49
Estel_so, you don't remove cable, etc, just battery09:49
luke-jr09:49
Estel_after that, You may remove cable if You feel it fancy, then You feed flasher with COMBINED09:49
kwtm2Well, the wiki says "you MAY have to remove battery" but it doesn't say you should, and previously I didn't remove battery.  So it's not so much that I don't understand the wiki as that there is conflicting info.09:49
luke-jrif you don't remove the cable, removing the battery doesn't really do anything09:49
luke-jrin my experience09:50
Estel_luke-jr, true, i just said to remove batteyr first. On second though, it doesn't matter09:50
Estel_kwtm2, it's because you may disable N90 0by other means than removing battery09:50
Estel_but it's said in wiki, that You must not allow it to reboot09:50
Estel_after eMMC flash09:50
Estel_anyway09:50
Estel_then Yo0u feed flasher with09:50
Estel_flasher-3.5 -F RX-51_2009SE_20.2010.36-2_PR_COMBINED_MR0_ARM.bin -f -R09:51
Estel_just watch uper and lwoer case f, i may have mixed them09:51
Estel_againt, first one is file09:51
kwtm2I see.  So after eMMC flash I *must* disable N900 somehow, e.g. removing the battery... I cannot just continue automatically with flashing firmware.09:51
Estel_then, "flash" command09:51
Estel_and last is reboot09:51
Estel_You may ommit the -r command, and reboot Yourself, but it's more conveinent09:51
Estel_yes.09:51
kwtm2Estel_: I do have the commands in a bash script that I wisely wrote when I first learned to flash.09:51
Estel_if N900 reboots between eMMC and COMBINED, your flashing gone fubar.09:52
kwtm2Ok, another thing that is unclear is... how do I reboot?  the power button doesn't do anything, and I don't know what's supposed to happen.09:52
Estel_nothing irrevertible - You just need to start from scratch09:52
Estel_when do You want to reboot? After Combined?09:52
kwtm2Ok, I will redo the flashing, removing the battery in between.  In answer to your last question,09:52
kwtm2I plan to flash both eMMC and firmware, so I guess I'll do that in that order.  But after firmware (last time) I tried to reboot but the device wouldn't respond.09:53
kwtm2So finally I took out the battery for a moment, replaced it, and it said "missing image".  So anyway, I will redo it the way you have told me.09:53
Estel_kwtm2, You may reboot using flasher - either provide -R at end of flashing command (combined only, not after eMMC0, or just do flasher-3.5 -R after flashing09:53
Estel_because it went fubar after eMMC flashing :)09:53
Estel_basically, you flash eMMC, without -R command to not allow it reboot - after flashing, You have quite unlimtied ammount of time to remove battery and unplug cable, no need to haste.09:54
kwtm2Estel_: Right, I've put -R now but previously I did not put -R and I thought I could reboot manually, which the wiki leads me to believe, but I see that this is not true.09:54
Estel_then, You flash combined, via same way as before (plugging device while holding "U"), but this time you may add -R command at end09:54
kwtm2Estel_: I did not reboot after eMMC, so I don't know why it went fubar.09:54
Estel_kwtm2, You *may* reboot manuall,y by issuing "flasher-3.5 -R" after flashing09:55
Estel_it';s just dividing it to two steps, unnecessary. More conveinent is to reboot after combined.09:55
kwtm2Estel_: ?? Do I need to plug the device in again while holding U?  That was not clear -- I thought I flashed the firmware immediately last time and it worked.09:55
Estel_theoretically, removing batteyr again, inserting it, and booting normally shgould work too, haven't checked09:55
kwtm2Ah, I reboot manually by issuing "flasher-3.5 -R" from my laptop.  That part is also unclear.  Ok.09:55
Estel_You flashed eMMC09:55
Estel_after eMMc, you remove battery AND unplug cable, as luke mentioned09:55
Estel_so you need to somehow plug it again to flash combined ;)09:56
Estel_and, you need to bring it into flashing mode agasin09:56
Estel_so, holding "U"09:56
Estel_basically, to flash including eMMc, You do two separate flashing, separated by removing battery. To flash normally, you just flash combined.09:56
Estel_warning - if You have done custom partition layout, it will be gone after flashing eMMC vanilla.09:57
kwtm2Ah.  That part was completely unclear.  There were 2 sections in the wiki, and they sounded independent, like written for people who only wanted to flash one or the other.  It was not clear that the U part was still necessary and could not be skipped if I wanted to do both.  Ok, I will put that in my script file.09:57
kwtm2Otherwise in 2 years I will forget again.09:57
*** sulx has joined #maemo09:58
Estel_kwtm2, thats true, wiki was written by different people.09:59
Estel_You could also upgrade wiki09:59
Estel_making it better source of knowledge not only for You, for later, but for others too09:59
Estel_but please, do it only after successful flashing :)09:59
kwtm2Estel_: Well, I successfully flashed about half a dozen times when I first got the device, two years ago, and wished I had updated the wiki then. :P10:01
Estel_so do it this timer, cause there is high chance that You will be using Yuor N900 after next 2 years too :)10:02
Estel_sorry for typos.10:02
*** vblazquez has joined #maemo10:02
kwtm2Estel_: Yes, I probably will; seeing Nokia's sinking ship, I bought an extra N900 because I don't want to give up my investment in all the scripts I've written for the N900 (and I don't want to have to learn Android or Apple).10:04
kwtm2That's partly why I dare to reflash my N900 now: I have an extra one in case something goes wrong.10:05
kwtm2The previous times it was out of necessity.  I learned firsthand about the ~/.profile bug. :P10:05
Estel_hehe.10:06
Estel_Yea, no really suitable devices on horizon10:06
kwtm2eMMC flashed.  So now I *must* not remove the USB cable when I remove the battery?  Or I *may* remove the USB cable before I remove teh battery?10:06
Estel_you may either buy android that got 21432453543543543543543 MHZ and RAMJ, just to spy on you more effectively, but don't allow You to use that horsepower...10:06
Estel_on anything usable...10:06
Estel_or stick to Your good N900 that does circles around multi-core devices, in terms of real life usability :)10:06
kwtm2I had such high hopes for The Samsung Linux Phone Secret Project.10:07
Estel_you may do it hoever you want10:07
kwtm2But I need to knwo.  Can I remove the cable before the battery?  Ah, I can.  Ok.10:07
Estel_just be sure to, at the end, have both batteyr and cable unplug10:07
kwtm2Ohh!!!  I see the firmware flashing now.  It did not do that previously, but listed all the stuff it was going to flash, and then exited to prompt, so I thought it had flashed already.  Someone should program the flasher to say "I did not flash"!!!  I thought the list of stuff was shown as it was flashing, and it was done.  Geez, no wonder.10:09
kwtm2Now I will have to go flash CSSU.  I want the power kernel that lets me do mobile hotspot.10:09
kwtm2Oh, and it's even rebooting.  Look at that.10:09
*** sulx has quit IRC10:10
kwtm2Holy cow, I must put in the wiki that it will flash does 5 dots for 67 seconds.  I was worried it had gone into infinite loop like the other time with the ~/.profile bug.10:12
*** ferdna has quit IRC10:14
kwtm2wait, the CSSU is just a package to be installed?  It does not need flashing?10:18
*** nmjnb has joined #maemo10:24
*** robbiethe1st has quit IRC10:26
kwtm2seen anyone10:26
Palikwtm2, yes CSSU is meta packge which install all updated packages10:26
*** sulx has joined #maemo10:26
kwtm2Ok, Pali.  Question about the power kernel (boy the web page is very ambiguous).10:27
*** mase76 has quit IRC10:27
*** vblazquez has quit IRC10:28
kwtm2I'm reading   wiki.maemo.org/Kernel_Power    It says with the stable kernel you can change the frequency limits, the voltages and DSP frequencies online."  Does that mean I *cannot* do the mobile hotspot (tethering)?  Do I need the unstable kernel for that?10:28
*** vblazquez has joined #maemo10:28
kwtm2It says of the unstable kernel "This kernel version contains additional experimental features and patches. "  But it doesn't say which additional features.10:28
*** utanapischti has quit IRC10:28
Palikwtm2, kernel-power is stable now10:29
Paliv5010:29
*** utanapischti has joined #maemo10:29
kwtm2The page "http://maemo.org/downloads/product/Maemo5/kernel-power-flasher/" lists features.10:29
Palilast version (v50) which is in Extras (and also on that link) is stable10:30
kwtm2Pali: Oh, so there is only one version of power kernel that I need to install, and it is both stable and contains the ability for hotspot?10:30
Paliyes10:30
kwtm2Also, it does not list hotspot but it lists NAT.  Is NAT what is required for hotspot?10:30
Paliyes10:30
Pali~wiki NAT10:30
kwtm2Ok.  Hmm, someone (probably me) should sort through that wiki and update it.10:30
infobotAt http://en.wikipedia.org/wiki/NAT (URL), Wikipedia explains: "{{Redirect|NAT||Nat (disambiguation)}} In computer networking, 'network address translation' ('NAT') is the process of modifying IP address information in IP packet headers while in transit across a traffic routing device. The simplest type of NAT provides a one-to-one translation of IP addresses. RFC 2663 refers to this type of NAT as 'basic NAT'. It is often also referred to as10:30
kwtm2I know what NAT is.10:31
kwtm2I did not know that that is what is required to run hotspot.  I thought all Linux came with NAT.10:31
kwtm2Anyway, now I know.  Will install CSSU and power kernel.10:31
kwtm2Thx.10:31
Paliyou can disable it when building linux kernel10:31
kwtm2Pali: Right, but I never figured Nokia would actually disable it.  it almost kills the point of having Linux.  Well, I am glad I can reenable it.10:32
Estel_kwtm2, wiki says about long time for 5 dots10:35
Estel_i remember, that it's written somewhere, about long time of first run due to optification10:35
kwtm2Estel_: Yes, but I think people would benefit from knowing a more absolute number.  15 seconds seems long to me (after all, the clock ticks at 600 million times a second, you'd think it would be done in 5 seconds).10:36
Estel_Pali, I remember now, thatone aprt of my tonight's dream was about forwarding info to You about osso-backup source codes being released10:36
Estel_I need vacation, I think10:36
Estel_kwtm2, there is no absolute number, it may be longer or shorter10:37
Estel_generally, it's just longer than usual10:37
*** LaoLang_cool has quit IRC10:40
*** Darkchaos has joined #maemo10:41
*** vblazquez has quit IRC10:43
kwtm2Wow, does the N900 change its mac address after reflashing!?10:44
kwtm2incredible!  So the firmware must have something that generates a MAC address after it's reflashed.10:45
Palikwtm2, no. mac address is stored in CAL NAND (/dev/mtd1)10:50
Paliand maemo system read it from nand at startup and send address to wifi chip10:50
kwtm2well, my n900 just got a new MAC address.  So maybe the nand is updated?10:50
*** Mrick0363 has quit IRC10:51
Palikwtm2, it could be corrupted10:51
Palior some app removed/changed your mac address in nand10:51
*** Mrick0363 has joined #maemo10:51
Palican somebody forward maemo.org website bugs to X-Fade?10:52
Palibug 1258110:52
povbot_Bug https://bugs.maemo.org/12581 Not possible to browse history of Developer forum10:52
Palibug 1262310:52
povbot_Bug https://bugs.maemo.org/12623 Do not auto log out on maemo.org midgard10:52
Palibug 1263310:52
povbot_Bug https://bugs.maemo.org/12633 Maemo.org website via HTTPS not working for anonymous users10:52
PaliEstel_ ^^^10:52
kwtm2Pali: It could be.  I have a record here that my N900 MAC address changed on 2010-06-19, so it's happened before.10:53
Pali2010?10:53
Palialso *original* mac address is somewhere on n900 box10:53
*** LaoLang_cool has joined #maemo10:56
kwtm2Really?  I don't know if I still have the original box.  In any case ... it changed again.10:57
*** ALoGeNo has quit IRC10:57
kwtm2Not that I mind.10:57
*** LaoLang_cool has quit IRC10:59
*** ALoGeNo has joined #maemo11:01
*** ALoGeNo has joined #maemo11:01
*** LaoLang_cool has joined #maemo11:03
*** LaoLang_cool has quit IRC11:03
*** Pali has quit IRC11:03
*** Pali has joined #maemo11:05
*** nsuffys has joined #maemo11:06
*** LaoLang_cool has joined #maemo11:10
*** mase76 has joined #maemo11:19
*** hurbu has joined #maemo11:25
*** e-yes has joined #maemo11:30
*** vblazquez has joined #maemo11:41
*** marainein has joined #maemo11:43
*** _federico3 has joined #maemo11:49
*** uen| is now known as uen11:55
*** robink has quit IRC11:57
*** sq-one has joined #maemo11:58
*** NIN101 has joined #maemo12:12
*** netkat has joined #maemo12:13
*** florian has joined #maemo12:13
*** OkropNick has joined #maemo12:14
*** Pali has quit IRC12:16
*** kerio92 has joined #maemo12:16
*** Pali has joined #maemo12:16
*** M4rtinK has joined #maemo12:18
*** kerio has quit IRC12:19
*** kerio92 is now known as kerio12:19
*** zeq has joined #maemo12:21
*** robink_ has joined #maemo12:22
*** vblazquez has quit IRC12:22
*** robink_ is now known as robink12:37
*** robink has joined #maemo12:37
*** leochenc1pher has quit IRC12:40
*** Guest17580 is now known as Jade12:45
*** Jade has joined #maemo12:46
*** Dr_I has quit IRC13:01
*** Cor-Ai has quit IRC13:07
*** rZr has quit IRC13:07
*** Pali has quit IRC13:12
*** Pali has joined #maemo13:15
*** eMHa_ has quit IRC13:18
*** eMHa has joined #maemo13:21
*** jhb has joined #maemo13:24
*** eijk has joined #maemo13:24
DocScrutinizerMacer: (bulging caps) you MUST use low-ESR types for replacement, usually13:29
*** kakashi__ has joined #maemo13:32
DocScrutinizerEstel_: yes, you're absolutely right about possibility of council asking x-fade (or - as group contact employer - Nokia in general) to change anything in IRC. I'm sure x-fade will follow _reasonable_ requests, to the degree his company policy allows (e.g. he mustn't give up control over channel, even if he maybe wasn't averse to do ;-D )13:34
DocScrutinizerEstel_: still I don't feel I pivot rules13:34
DocScrutinizerEstel_: let's agree that we disagree sometimes about how to read any existing rules13:36
freemangordon zeq: trying GTK build with HW acceleration enabled, it was a mistake (or better said ignorance :) ) by my side when I said the enabling GLES with GTK does not make any difference. There is about:config value that defaults ot"HW acc disabled", but I find it after i've tried GTK13:36
freemangordonGTK build uses 10-20 less MB, I think that is the option we have to find a way to use13:36
SpeedEvilfreemangordon: Of real memory - PSS ?13:37
freemangordonyep, RAM, according to conky13:37
SpeedEvilThat's misleading, it's probably using RSS, which has issues13:37
SpeedEvilyou don't care about - for example - the amount of RAM used for mapped libraries that are mapped by system processes13:38
freemangordonSpeedEvil: well, Qt is not mapped ;)13:38
freemangordon(libs)13:38
SpeedEvilk13:38
freemangordonwhile GTK is13:38
SpeedEvilI just woke up, so I may be say ing crap.13:38
freemangordon:)13:38
SpeedEvilhttp://www.selenic.com/smem/13:38
SpeedEvilWorks fine on n900 - the python bit, not the graph13:38
SpeedEvilI found it useful.13:39
freemangordoneven with thumb, newer compiler, etc,etc, Qt is still HUGE13:39
SpeedEvilSome of the trivil nokia resident stuff is _very_ lightweight13:39
SpeedEviltrivial13:39
freemangordonand if we find a way to get rid of it for fennec(which itself is a memory hungry animal), we will gain some speedup,whatever13:40
*** valeriusN has left #maemo13:40
freemangordonalso i tried flash with fennec, it works, but it is SLOOOW. And I can bet it is because of Qt13:40
SpeedEvilFlash is crap, and should die. But it's very useful :/13:41
freemangordonback in the days I was tring to make my own Qt/WebKit based browser, it was PITA to run flash without performance penalty because of Qt13:41
freemangordonNot that I did not succeed, but it was PITA13:42
freemangordonanyway, lets see how GTK will behave with GLES acceleration enabled13:42
*** jhb has quit IRC13:43
*** GeneralAntilles1 has joined #maemo13:43
*** GeneralAntilles has quit IRC13:43
DocScrutinizerEstel_: (( If You think that referendum can be used to remove volunteers form volunteering,...)) please reread my post - I *never* suggested _anything_ like that. Such a referendum would be outright silly, unless it was about massive (criminal) lapse of one particular council member, in which case I guess most democracies know the concept of removing immunity from politicians. In same context: I think applying rules known from good13:44
DocScrutinizerdemocratic constitutions / politics is the only thing we got for most "common sense" questions regarding council and maemo 'administration' at large13:44
freemangordonSpeedEvil: in the meantime I agree with you re flash. But on the other hand if we have a working plugin on n900, why not use it?13:44
SpeedEvilTotally agree.13:44
*** FredrIQ has joined #maemo13:45
DocScrutinizerEstel_: the basic concept however is always same, no matter how silly the request: first you collect supporters for your idea; if you can find sufficient number of supporters it will get attention of the administrative entity which then decides (based on whatever rules) if they will start a referendum about that topic13:47
*** FIQ has quit IRC13:48
*** FredrIQ is now known as FIQ13:48
DocScrutinizerit's one of the most honourable duties of the administrative entity to make sure common sense applies to any such request and that no reasonable well supported request goes unheard13:49
DocScrutinizerchanging rules of council elections is a reasonable request in my book, if that change cures a perceived flaw in the present rules13:50
*** MohammadAG has quit IRC13:53
*** xnt14 has quit IRC13:54
*** valerius has quit IRC13:54
*** nsuffys has quit IRC13:54
Siceloso MfE is one hell of a beast on N900. m.google.com for server does not put any packet at all on outgoing pipe. it's like this thing is set to ignore m.google.com. otoh, its IP address works, but gets stuck because the certificate only mentions m.google.com13:55
freemangordonSicelo: do you have newer libcurl3 installed?13:55
Siceloi had older one yesterday. same thing. then i upgraded, same thing too13:56
*** brzys has joined #maemo13:56
freemangordonno offence, which libcurl3 version you have now?13:56
freemangordonit should be 7.25.0-1maemo713:57
Sicelo7.25.0-1maemo713:57
freemangordonSicelo:^^^13:57
freemangordonok13:57
Siceloyeha13:57
*** dos1 has joined #maemo13:59
Sicelofreemangordon: this is a bad one?13:59
freemangordonit is good14:00
Sicelojust interesting why MfE doesn't send anything at all. i would understand if it sends wrong or incompatible stuff. but here, nothing at all (according to wireshark)14:01
freemangordonSicelo: did you restart after installing libcurl3?14:01
Siceloany other url/ip does at least result in some sent packets, except m.google.com14:01
Sicelohmm, i don't remember. let me do it just in case14:02
freemangordon:nod:14:02
*** muelli has joined #maemo14:02
*** Neutron_ has quit IRC14:03
*** Neutron_ has joined #maemo14:04
DocScrutinizermoo folks, btw14:04
*** maybeArgh has joined #maemo14:05
*** RiD has joined #maemo14:06
Sicelolulz, works after a reboot. thanks freemangordon :P14:07
freemangordon:P14:07
*** maybeWTF has quit IRC14:09
Siceloi'm still puzzled why it exhibits this behavior14:10
freemangordonDNS type AAAA records14:12
*** Atarii has joined #maemo14:16
*** Atarii has quit IRC14:16
*** Atarii has joined #maemo14:16
*** valerius has joined #maemo14:16
*** mase76 has quit IRC14:17
*** mase76 has joined #maemo14:18
*** FIQ|n900 has quit IRC14:18
*** dhbiker has quit IRC14:25
DocScrutinizerSicelo: :o)14:29
*** psychologe has joined #maemo14:29
* DocScrutinizer send Sicelo off for 6 moths windows training, to learn about "you changed position of mouse - please reboot to make the changes to pointer take effect"14:30
*** erkules|away has quit IRC14:31
*** erkules|away has joined #maemo14:31
*** erkules|away is now known as erkules14:31
Sicelohmm,:P14:31
Sicelolulz.14:32
*** ced117 has joined #maemo14:32
*** ced117 has quit IRC14:32
*** ced117 has joined #maemo14:32
*** vblazquez has joined #maemo14:34
DocScrutinizeranyway, freemangordon I guess we *need* a mandatory system reboot in libcurl postinst14:34
DocScrutinizermaybe a tiny bit of overkill (since in RL a restart of all services using libcurl should probably suffice), but MEH14:35
Siceloi know i did reboot, but i suppose now that it was before this libcurl update. that said, i guess i should have known upgrading a lib would need a reboot14:36
freemangordonDocScrutinizer: anyway, libcurl3 is not supposed to be distributed through extras-devel14:36
DocScrutinizerLOL right14:36
DocScrutinizerwhat's the name of the friggin service that waits til apt-worker and HAM and whatnot all have quit, then notifies user "now boot is pending! you can't dismiss this requester!"14:38
DocScrutinizer(assuming there has to be such process/script/tool already)14:39
Estel_DocScrutinizer,  so in fact, we completely agree.14:42
Estel_You were describing process for "any" big change14:43
*** valerius has quit IRC14:43
Estel_which, of course, fall for referendum14:43
Estel_falls under*14:43
Estel_in context of person asking question etc. it could be, hoever, interpret as Your suggestion, how he shoul.d proceed ikn *his* case14:44
*** sq-one has quit IRC14:44
Estel_i.e. someone not satisfied by CA results -> referendum about removing councilor(s). That was why I called it silly14:44
Estel_I think that You mean procedure as whole, generaly14:44
jacekowskiremove CA14:44
jacekowskiproblem solved14:44
Estel_in which case, as said, we totaly agree14:44
Estel_jacekowski, ca is removed, it's ended :P14:45
Estel_DocScrutinizer,  just keep in mind that I wasn't only one who took Your answer there as invitation to make silly trollstorm referendum - it seems that lma got same feeling, and, possibly, other readers14:45
Estel_just problem of misscomunication and context :)14:45
Estel_freemangordon,  ping14:46
freemangordonpong14:46
Estel_about cssu-thumb14:46
Estel_(if You have time now)14:46
freemangordonwhat about it?14:46
Estel_what are plans for painless integration with current, erm, ecosystem (joking :P) of programs depending on KP?14:47
Estel_i.e. will kernel-cssu "provide" kp in future?14:47
Estel_I feel silly changing control files and repacking everything that want to see KP and have kernel-cssu instead14:47
Estel_...from kernel-power-settings, to cleven and h-e-n14:47
freemangordonso far there are 2 packages which don't like kernal-cssu - KPS and compat-wireless-power14:47
Estel_disclaimer - I'm not demanding or whinning, just curious14:48
*** magog has joined #maemo14:48
freemangordonh-e-n has no problem14:48
Estel_ough, so I did it without reason14:48
Estel_so it's rather packages in question problem, not kernel-cssu one?14:48
freemangordonPali, ping14:48
Palifreemangordon, pong14:48
freemangordonEstel_: yep14:48
freemangordonPali: will you push newer KPS?14:49
Estel_also, is there anything depending on cssu (normal, not thumb) that cries when thumb one is present?14:49
Estel_Pali, I've seen bugs about web, forwarded them14:49
freemangordonnone I am aware of14:49
freemangordonPali: maybe new KP too14:49
PaliEstel_, ok14:49
Estel_Pali, yes, please, newer KPS would be greatly appreciated - as I presume, it will work with both KP and KCSSU with no problems14:49
freemangordonwith the fix for FMTX14:49
Estel_from now on14:49
Estel_wut fix for ftmx?14:50
Palifreemangordon, there was problem with kernel-power-settings some shift problem when user has old config file14:50
freemangordonPali: I know14:50
Palido you have log?14:50
PaliI'd like to fix this too14:50
PaliEstel_, echo 1 > lock for fmtx not worked14:51
freemangordonPali: no14:51
Estel_I see14:51
Palibut printf 1 > lock worked (e.g omit \n)14:51
Estel_erm, what this echo was supposed to do?14:51
Palilock or unlock fmtx14:51
freemangordon(actually it causes sh to use 100% CPU)14:51
Estel_ouh14:51
Estel_understood14:51
Estel_kernel-cssu also contain fix?14:51
freemangordonno :)14:52
Hurrianfreemangordon, it seems that linux-backports-modules-power depends on KP too14:52
freemangordonHurrian: yep14:52
freemangordonand guess who is the maintainer :D:D:D14:52
*** Estel_ is now known as dhfg14:52
dhfg~span freemangordon14:52
dhfg~spank freemangordon14:52
* infobot bends freemangordon over his knee and tatoos 'ibot' on freemangordon's pasty white buttocks.14:52
*** croppa has quit IRC14:52
Hurrianit should probably be inlined into the kernel now14:52
*** dhfg is now known as Estel_14:52
Estel_freemangordon,  seriously though, to this day I wonder, what is use for those bluetooth modules?14:53
Estel_compat one?14:53
Estel_what use is*14:53
Hurrianwhelp, ran out of time this weekend, but come next weekend, I'll try winging it and simply replacing the driver trees and rebuilding14:53
freemangordonHurrian: a better solution is Provides: kernel-feature-wireless-compat-VNN14:53
freemangordonPali: ^^^14:53
Estel_agree.14:53
freemangordonwhat do you think14:53
Hurrianfreemangordon, agreed14:54
DocScrutinizerEstel_: (problem of context) quite likely. I strictly answered on this one post of zogg - not even reading the prev thread14:54
freemangordonEstel_: NFC what these BT modules do, better as whoever was the author of wireless-compat drivers14:54
freemangordoni've just packaged it14:55
Estel_DocScrutinizer,  this answers it. Of course it's not Your fault, but You may consider (or not) some clarification, as, per what I know, 100% of readers took it as support for zog's idea14:55
Hurrianalso, imo kp-based kernels should provide kernel-feature-poweruser or something to indicate presence of augmentations14:55
Estel_freemangordon,  aye. So we're not using them anywhere? I wonder if anyone tried loading them14:55
Estel_and live to tell tale about effects14:55
freemangordonEstel_: NFC. You're free to try them :P14:55
zeqfreemangordon: I'll be about later this afternoon.  I agree, we need to go with whatever works best qt vs gtk, and memory consumption is a big one!14:56
Estel_GTK! GTK! just for the sake of it :P14:56
freemangordonzeq: yep, unfortunaltely GTK build spits "Attempting load of libEGL.so14:56
freemangordonFailed to create EGL config!"14:56
PaliHurrian, see u-boot thread for initrd14:56
zeqdoes it fail to find libEGL.so?14:57
freemangordonit fails to create EGL config14:57
Palifreemangordon, where to add Provides?14:57
Estel_DocScrutinizer,  stupid question mode on - considering N900 board layout, are there any chances for hassle-free inclusion of active cooling, as prrof-of-concept? Or due to sandwitched chips, it would be placebo?14:57
* freemangordon reboots the device in an attempt to make that work14:57
Palito package linux-backports-modules-power?14:58
freemangordonPali: in KP14:58
zeqfreemangordon: similar to the situation with qt with egl enabled?14:58
freemangordonzeq: no14:58
Hurrianzeq: use one, and only one14:58
freemangordonQt port works14:58
SpeedEvilEstel_: Cooling is largely pointless14:58
Estel_SpeedEvil,  hm?14:58
SpeedEvilEstel_: Issues with overclocking are not due to temperature.14:58
Estel_of course14:58
freemangordonPali: right now compat-wireless-power depend on KP ==== v5014:58
freemangordon==14:59
Estel_I'm not telling, that it will create devices with 1.7 GHZ14:59
zeqfreemangordon: we'll need to look into where all the memory is going14:59
kwtm2Just woke up. Sorry, there is now a kernel-cssu which is not kernel-power? If I want mobile hotspot, do I need kernel-power or is kernel-CSSU enough?  I dont' need to overclocking; I just want mobile tethering for my portable wifi devices like the non-cellphone-enabled iPad we have.14:59
Estel_SpeedEvil,  I'm thinking about it as attempt for counter-electromigration14:59
Hurriankwtm2, kernel-cssu is kernel-power14:59
Estel_prolonging live of device14:59
freemangordonzeq: it is Qt by itself14:59
Hurrianjust... rebranded14:59
Estel_and, generally, cooler cpus works better :P14:59
DocScrutinizerEstel_: (cooling) terribly complicated and of rather questionable benefit14:59
*** hurbu has quit IRC15:00
*** mase76 has quit IRC15:00
DocScrutinizerahh, cheers SpeedEvil15:00
zeqfreemangordon: so all other qt apps so similar usage?15:00
Palifreemangordon, ok, so you need some provide line kernel-power-flasher and kernel-power-bootimg15:00
freemangordonyep, even worse without thumb15:00
Estel_i.e. applying radiator to part of SoC being most top one isn't going to help?15:00
HurrianEstel, the most we're putting into the N900 is a small heatsink15:00
kwtm2I see. I must have misunderstood  [04:47] <Estel_> I feel silly changing control files and repacking everything that want to see KP and have kernel-cssu instead15:00
Estel_what it is, btw, one-nand?15:00
SpeedEvilEstel_: You are putting stress on the ram-chip15:00
kwtm2If they are the same, then that's ok.  In fact, I fell asleep trying to upgrade the kernel on my N900.15:01
Hurrianthe onenand sits on top of the CPU15:01
SpeedEvilEstel_: which has issues of its own. Cooling the board is probably better.15:01
DocScrutinizerEstel_: regarding EM it's pointless since due to very local and sub-millisecond hotspots on chip you physically can not transport temperature away fast enough15:01
Estel_Hurrian,  sure, heatsink sounds reasonable...15:01
*** hurbu has joined #maemo15:01
Estel_DocScrutinizer,  understood. General chip temperature, prolonged, doesn't affect it, though?15:01
DocScrutinizerit would when you could cool down the chip by 100 or 150 200 °C15:02
Estel_SpeedEvil,  cooling the board? using what? liquid nitrogen? :P15:02
Hurrianon the more extreme side, with the addition of another battery, you may be able to strap on a peltier cooler15:02
Estel_Hurrian,  peltier would introduce more heat15:02
Estel_to take it out15:02
Estel_+ isolation15:02
DocScrutinizercooling the chip means cooling the whole board as the board is chip's heatsink15:02
HurrianEstel, we just have to drain it fast enough, somehow15:03
*** hurbu has quit IRC15:03
Hurrianagain, impractical but cool15:03
DocScrutinizerbottom line: interesting idea - but forget it ;-)15:03
Estel_Hurrian,  more 1337 is using 75 percent potasium and 25 percent Na - melting point 25 celsius degrees under zero. Call it liquid metal cooling :P15:03
Estel_DocScrutinizer,  lol15:04
*** hurbu has joined #maemo15:04
Estel_SpeedEvil,  so generally applying heatsink to EMI shields (as they're connected to mass, ground plates inside board) is as much practical as fancy cooling of chip itself?15:05
Hurrianmaking a cooler for the N900 wouldn't be all that hard, there's a lot of free space for a milled/stamped out passive heatsink15:05
Estel_btw, emi shields are not soldered, but simply "put" on borders, borders itself are soldered15:05
Estel_Hurrian,  yep.15:05
Hurrianoptionally an "active" heatsink, with the addition of blowing air15:05
DocScrutinizerHurrian: won't help anything, as the bottleneck is the case itself15:05
Estel_I got just crazy ideas for my first prototype of aluminum body replacement :P15:05
Estel_some of them containing heat pipes15:06
*** hurbu has quit IRC15:06
Estel_:P15:06
*** valerius has joined #maemo15:06
Estel_I know, silly, I was thinking about it as a proof of concept, rather15:06
HurrianI have a salvaged copper laptop cooler on my table, and I was imagining all morning how easy it'd be to stamp one out, and thermal-tape it to the SoC15:06
*** hurbu has joined #maemo15:06
DocScrutinizerEstel_: with an alu body the whole thing changes a LOT15:07
*** psychologe has quit IRC15:07
Estel_well, I wonder how cooling one-nand (being the topmost part) would propagate to SoC and rest of noard15:07
Estel_DocScrutinizer,  yea15:07
DocScrutinizernot at all15:07
HurrianEstel_, just regular heat propagation, but slower15:07
Estel_I would like to most effectively transfer heat from board/chip/memory to alu body15:08
DocScrutinizerSoC has power eating CPU in sandwich between board and RAM POP15:08
Estel_and I'm searching for best "contact point" or points15:08
Estel_yea15:08
DocScrutinizercooling RAM POP won't earn you much15:08
Hurrianalu/copper sleeve that wraps around the entire SoC+PoP assembly?15:08
Estel_wait, so ram is topmost, not one-nand?15:08
DocScrutinizeryou need to cool the board15:08
DocScrutinizerground plane of board to be specific15:09
Estel_DocScrutinizer,  yea, suspected that15:09
Estel_generally, they did awesomely good job of propagating heat transfer to board. Now how to take it out effectively, without big direct contact point with ground plane?15:09
Estel_emi shields are just "put" at top, not soldered15:10
HurrianEstel_, encase the whole SoC and PoP in a sleeve?15:10
DocScrutinizersince there ate few plain void areas on board, you have a hard time to physically attach a heatsink to it15:10
Estel_Hurrian,  it is some idea, but I'm rather thinking about taking it out from shields inside board (out heat)15:10
Hurrianah, without taking off the shields?15:10
Estel_maybe scrap solder mask somewhere? :P15:10
Estel_Hurrian,  yea, or with, as my alu body is shield in itself15:11
Estel_so it matters less15:11
DocScrutinizerEMI shields don'T help - it's iron, very thin iron. A rather poor heat conductor15:11
*** klink has joined #maemo15:11
Estel_exactly15:11
Estel_+ it is in just "touch" contact15:11
Estel_first idea was to put heatsink on chip itself, attached via liquid metal as a thermal paste (seriously), it's easy doable, but rather poor at cooling board15:12
Estel_well, helps a little due to stressed ram being cooled... but...15:13
DocScrutinizeryou might (given the bizarre nature of whole topic) consider sealing SoC shield can hermetically, then attach two nozzles and pump cooling liquid thru the whole compartment15:13
Estel_0_o15:13
*** Termana has joined #maemo15:13
DocScrutinizer:-D15:14
Estel_N900 underwater project :D15:14
DocScrutinizerprobably some special mineral oils would work15:14
DocScrutinizeror some high alcohols15:14
Estel_yea, but they're bad as coolers :P15:14
HurrianEstel_, mineral oil board immersion?15:14
DocScrutinizernah, glycol is not that bad15:14
Estel_Hurrian,  yea :P15:14
HurrianI can see TMO filled with requests for a special fiberglass mineral oil case15:15
DocScrutinizerbut I guess it's too hygroscopic to be a good isolator15:15
Estel_we have some points where keyboard case and such things touch ground planes15:16
*** mase76 has joined #maemo15:16
Estel_or generally, parts of body15:16
Estel_putting them in contact with alu body - in good contact - is probably going to help a lot15:16
DocScrutinizerrealistically your best bet is moving air across the 'backside' of PCB15:16
Estel_Ion cooler :D15:17
Estel_2KV in Your N900's backside15:17
DocScrutinizerwhy not15:17
Hurrianwe should do that without cutting a large hole in the phone and decreasing resistance to water drops15:17
Estel_would be cool, and could dubble for a neon light power source15:17
*** shanttu has joined #maemo15:18
DocScrutinizerotoh ther are extremely small fans available or you can build them from RC-components15:18
Estel_I think that, realistically, touching ground plates where applicable is best idea, + part touching chip without emi shielding as a extreme case15:19
Estel_huh?15:19
Estel_power suckers :P15:19
DocScrutinizera bit15:19
DocScrutinizernot that bad15:19
DocScrutinizerway less than the energy they move out ;-P15:19
Estel_we were joking a little, but seriously I'm thinking now about realistical ways to increase cooling, having alu body15:19
Estel_DocScrutinizer,  yea :D15:20
Estel_BTW, You know that someone did thievery on my idea and is selling phone chargers based on flame and heat?15:20
DocScrutinizerfind plain void areas of PCB, glue headsinks/-pipes to them with CPU cooler sticky pads15:20
Estel_DocScrutinizer,  sticky pads are suckers at effective cooling, I'm using liquid metal instead :P but generally, agree15:21
*** dhbiker has joined #maemo15:21
Estel_my idea was stirling engine charger powered by pocket lighter15:21
Estel_someone decided, that incorporating room where You put a little wood and burn it is better, lol15:21
DocScrutinizerliquid metal doesn't isolate, a mandatory property of your heat conducting stuff you use to attach to PCB geberally15:21
Estel_DocScrutinizer,  but I'm attaching metal part to ground, why should it isolate?15:22
Estel_current N900 case is grounded too15:22
Estel_backside of screen, metal frame around it15:22
DocScrutinizerdon't mess with groundplane design!15:22
DocScrutinizerit's not a static thing, electrically15:23
Estel_? why, it's interesting15:23
DocScrutinizerany loop you create may cause massive trouble15:23
DocScrutinizerwhole groundplane is also part of all antennae15:23
Estel_I see, so why they have, practicaly, attached back of screen and our metal loop around screen to ground plane? or i missunderstood something?15:23
DocScrutinizerthey evaluated this particular way they did this, and you mustn't change it basically15:24
Estel_sure, I'm just interested why, natural curious. BTW, I got electricized by N900 frame around screen, once15:25
*** valerius has quit IRC15:25
DocScrutinizergeneral rule: keep any existing ground connections, don't create any new ones15:25
Estel_my modified charger was screwde in one point, and applied AC voltage to USB case15:25
*** klink_ has joined #maemo15:26
Estel_harging worked fine, but voltage from outer part of USB case (not 220V, lower) was transfered to N900's usb case15:26
Estel_and through it, to metal frame around screen15:26
Estel_nice anti-thievery protection :P15:26
*** vblazquez has quit IRC15:27
DocScrutinizerI heard of somebody getting this effect on headset earplugs X-P15:27
Estel_it was lying on my desk, and every time I reached my notebook (which have metal case), something was biting my arm15:27
Estel_lol15:27
Estel_then I started to suspect something - for first, I though that my notebook metal case is on voltage15:27
Estel_started to emasure it with multimeter15:28
DocScrutinizerTHIS must go directly to your brain! ;-P15:28
Estel_to the point where I touched notebook and N900 metal case via multimeter - notebook to this point have small brown point, burned out15:28
Estel_just on surface, but still15:28
*** hurbu has quit IRC15:28
DocScrutinizerEstel_: either you switched your MM to ampere, or you used a really really crappy MM15:29
Estel_Part of charger pcp cut through isolation while assembloing unit, and was touching one capacitor or whatsnot, anyway, it was creating initially small, yet getting bigger and bigger voltage15:29
Estel_the latter :P15:29
*** mase76 has quit IRC15:29
DocScrutinizerimpedance of a MM at voltage range should be several M-Ohms per volt15:29
Estel_exactly.15:29
Estel_which convinced me to buy new MM :P15:30
*** mase76 has joined #maemo15:30
DocScrutinizergood :-P15:30
Estel_anyway, using ampere in search for wandering voltage would grant me honorable mention at darwin awards15:30
*** klink_ has left #maemo15:30
Estel_btw I'm glad my device wasn't screwed due to this, fortunately, board is wisely designed15:31
Estel_(i.e. AC 40-60V applied to ground plane didn't affected N900at all)15:31
Estel_it was working whole time :P15:31
Estel_btw I wonder what the hell made voltage AC that was initially less than 1V and was raising with time15:32
kwtm2Sorry, how do we know/show that kernel-power is installed ?  (apart from apt-cache policy or otherwise apt *saying* it's installed.) Can we tell from uname in terminal?  Some settings screen that says "I now feel omnipotent" or something?15:33
*** vblazquez has joined #maemo15:34
*** klink has quit IRC15:34
*** klink has joined #maemo15:35
Sicelouname15:35
Estel_DocScrutinizer,  I waas thinking about this preserving ground plane design and something came to my mind:15:36
Estel_1. We know, that metal frame and backside of screen is connected to ground plane.15:36
*** muelli has quit IRC15:36
Estel_2. So, by touching metal frame to any metal, even accidentaly, we change groundplane design!15:37
*** hurbu has joined #maemo15:37
Estel_yep?15:37
*** retro|cz has quit IRC15:37
Estel_i.e. when N900 lies on desk, touching box for cigarettes (metal one) ground plane is different, yes?15:38
DocScrutinizerEstel_: better? http://talk.maemo.org/showpost.php?p=1229427&postcount=4615:45
shanttuFinnish technews-site is gathering information on Maemo/Meego/Meltemi -story. Author's first words on topic http://tinyurl.com/75ao493 Will be interesting to read the article when it's ready.15:45
*** mase76 has quit IRC15:46
DocScrutinizer(ground plane design) basically yes. Really nasty problems are created by loops though, which obviously need 2 touchpoints15:48
*** valerius has joined #maemo15:48
*** dhbiker has quit IRC15:49
*** dhbiker has joined #maemo15:51
*** dhbiker has quit IRC15:58
*** shanttu has quit IRC16:03
*** brzys is now known as QnJ6ecWb16:03
*** Evanescence has joined #maemo16:04
*** dhbiker has joined #maemo16:04
*** kwtm has joined #maemo16:05
*** kwtm2 has quit IRC16:05
*** disco_stu has quit IRC16:05
*** Evanescence has quit IRC16:07
Estel_DocScrutinizer,  sure (edit)16:07
*** stardiviner has joined #maemo16:07
*** stardiviner has joined #maemo16:08
*** sq-one has joined #maemo16:09
*** stardiviner has quit IRC16:10
*** jluisn has joined #maemo16:11
*** vblazquez has quit IRC16:13
*** cuonglb has joined #maemo16:13
*** cuonglb has quit IRC16:15
*** user_ has joined #maemo16:17
*** DocScrutinizer has quit IRC16:18
*** DocScrutinizer has joined #maemo16:18
*** user_ has quit IRC16:19
*** user_ has joined #maemo16:19
*** disco_stu has joined #maemo16:20
kwtmWow, the built-in backup manager on the N900 froze (on both of my N900's).  Wonder if it's a bad microSD card?16:20
*** vetsin_ has joined #maemo16:22
*** hurbu has quit IRC16:25
*** QnJ6ecWb has quit IRC16:26
*** hurbu has joined #maemo16:28
*** vetsin_ has quit IRC16:29
*** mase76 has joined #maemo16:30
*** Pali has quit IRC16:31
*** Zahra has joined #maemo16:31
*** Pali has joined #maemo16:33
*** Smusmumrik has joined #maemo16:33
*** ALoGeNo has quit IRC16:33
*** Smusmumrik has quit IRC16:35
DocScrutinizerkwtm: backup manager works on "plugins" for apps to tell what they want to get backup'ed and how. If such a plugin is broken by design, it might freeze whole backup manager I'd guess16:36
*** jluisn has quit IRC16:37
*** netkat has quit IRC16:37
*** Pali has quit IRC16:39
*** dhbiker has quit IRC16:40
kwtmDocScrutinizer: Ok. But I was just trying to restore from a reflashed clean system. It asked me what I wanted to restore (communications, apps list, etc.) and when I pressed ok on the default (which was all 4 of those items to be restored) it froze.  Wouldn't even let the sliding switch lock the screen/keys.16:44
*** dhbiker has joined #maemo16:44
kwtmPerhaps by that time it had already installed a broken plugin?  I don't know. Anyway, I removed the battery to reboot.16:44
*** Zahra has quit IRC16:47
*** hurbu has quit IRC16:47
*** hurbu has joined #maemo16:49
*** eijk has quit IRC16:49
*** hurbu has quit IRC16:51
*** hurbu has joined #maemo16:54
*** lxp1 has joined #maemo16:54
*** lxp has quit IRC16:56
*** shanttu has joined #maemo16:58
DocScrutinizerstrange thing is that my device yesterday also semi-froze, wasn't able to unlock screen anymore, though other stuff still worked evidently (I was able to tell from my playswoosh sound attached to kbd slide open event)16:58
DocScrutinizeralso had to remove battery to recover16:59
DocScrutinizersince not even 10 seconds press of powerbutton helped to shut it down gracefully17:00
DocScrutinizerI'm still wondering what caused this17:01
*** pauly has joined #maemo17:01
*** pauly is now known as Guest8560817:02
*** Guest85608 is now known as Pauly1217:02
ShadowJKDocScrutinizer; oh hey, I need your paypal address17:03
*** nmjnb has quit IRC17:07
*** e-yes has quit IRC17:08
*** hurbu has quit IRC17:11
*** florian has quit IRC17:16
ZogG_laptopShadowJK: i need your clothes sunglasses and motocycle17:16
*** nmjnb has joined #maemo17:17
ShadowJKhave a twinkie17:17
DocScrutinizerShadowJK: there's a "support my work" link under each tmo post of me17:18
Pauly12hello17:19
ShadowJKhow much did the n900 cost you?17:19
DocScrutinizerer, some 250..26017:20
DocScrutinizerlemme check17:20
DocScrutinizer25517:21
DocScrutinizerPauly12: let me guess - you got a N900 to sell?17:22
DocScrutinizer:-)17:22
Pauly12no17:22
DocScrutinizerooh sorry, then my memory failed17:23
Pauly12DocScrutinizer: wanted to know if you can take RAW pics on n9?17:23
DocScrutinizerI don't know17:23
DocScrutinizeranyway I guess #n9 or #harmattan are the better channels to ask that17:24
ZogG_laptopat least there is no setting for that17:24
Pauly12the Fcam api is in harmattan but i have no idea if there is an app17:24
Pauly12DocScrutinizer: ya i regret selling my n90017:26
*** hurbu has joined #maemo17:27
*** hurbu_ has joined #maemo17:31
*** vblazquez has joined #maemo17:32
*** GeneralAntilles1 is now known as GeneralAntilles17:35
*** GeneralAntilles has joined #maemo17:35
*** hurbu_ has joined #maemo17:35
*** hardaker has joined #maemo17:36
*** vblazquez has quit IRC17:39
*** freemangordon has quit IRC17:39
*** AsianFemale24 has joined #maemo17:40
AsianFemale24hello17:42
AsianFemale24hello?17:42
*** AsianFemale24 has left #maemo17:42
*** radic has quit IRC17:46
*** radic has joined #maemo17:48
*** hurbu_ has joined #maemo17:49
*** kerio has quit IRC17:49
*** mase76 has quit IRC17:49
*** SpeedEvil has quit IRC17:49
*** hurbu_ has quit IRC17:49
*** SpeedEvil has joined #maemo17:50
*** kerio has joined #maemo17:51
*** LaoLang_cool has quit IRC17:55
*** ArGGu^^ has quit IRC18:05
*** freemangordon has joined #maemo18:10
*** hardaker has quit IRC18:15
*** ArGGu^^ has joined #maemo18:24
freemangordonzeq: http://hg.mozilla.org/users/romaxa_gmail.com/embedipc_queue18:25
freemangordonhttps://wiki.mozilla.org/Embedding/IPCLiteAPI18:25
kwtmWow, when I reinstalled old apps from a list, I got a nice widget on my Maemo desktop screen that shows a phone dialpad where I can actually dial without going to the phone app.  How can I find out what widget this is called?18:28
*** e-yes has joined #maemo18:36
*** ArGGu^^ has quit IRC18:42
*** PeterWolf has quit IRC18:51
Sicelo~ugt18:52
infobotmethinks ugt is Universal Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. http://www.total-knowledge.com/~ilya/mips/ugt.html18:52
*** ArGGu^^ has joined #maemo18:55
DocScrutinizerdafaq, AsianFemale2418:57
DocScrutinizerLOL18:57
freemangordonhmm, sounds good, where did you find her :P18:58
DocScrutinizer10 lines up18:58
DocScrutinizerkwtm: try googling "maemo rotary dialer"18:59
freemangordonunfortunately by that time my windoze was slowly creeping south :D19:00
zeqfreemangordon: that sounds like what we want if we're aiming to replace microb19:00
freemangordonyep19:01
freemangordonlooks like WIP though19:01
kwtmDocScrutinizer: Thanks! But how do I, in general, find out what the widgets are on my desktop?  Is there a ~/.maemo/.desktop/*.widget sort of directory/file?19:01
DocScrutinizerfreemangordon: try direct nmap! ;-D [314] AsianFemale24 ~Admin 213.87.121.46 * purple purple19:01
zeqfreemangordon: I wonder how that relates to native fennec?19:02
freemangordonzeq: i finally managed to turn GTK EGL acc on, unfortunately it segfaults in libEGL.so, now trying without --enable-egl-xrender-composite19:02
freemangordonzeq: most probably not related19:02
freemangordonif I got your question right19:03
zeqfreemangordon: I tried to build qt with --enable-egl-xrender-composite, it didn't compile though... did you get that to work?19:03
freemangordonsame here, it does not compile19:03
zeqno just me then :)19:03
freemangordonbut GTK version compiles19:03
zeqs/no/not/19:03
infobotzeq meant: not just me then :)19:03
freemangordonunfortunatel Qt build is SLOOW with flash enabled :(19:04
freemangordonsome 1-2 fps19:04
zeqI read ^^^ you've some experience with that19:04
luke-jrfull Qt build is impossible on N900 natively :/19:04
freemangordonyeah, but i was using webkit19:04
freemangordonluke-jr: what?19:04
luke-jrQt-WebKit will lock the system swapping so hard that it watchdog-reboots19:05
freemangordonyou mean on the device19:05
luke-jryes19:05
zeqok... I wonder why it's slow... and how relevant it is any more...19:05
freemangordonit spits something about XEmbed19:05
freemangordonnot using shm or something19:05
freemangordonNFC19:05
zeqthat would do it19:05
zeqif it's having to copy pixmaps around all over the place19:06
freemangordonyep19:06
freemangordonbut that is strange, as webkit works pretty ok19:06
zeqmust be a bug19:06
freemangordon(with some hacks)19:06
*** shanttu has quit IRC19:06
zeqsomething with the plugin container19:06
zeqsomething up*19:07
freemangordonyeah, this new IPC is lots of problems AIUI19:07
freemangordonnot that we can fix it :D19:07
zeqlooking at the specs now19:07
freemangordonwhich specs?19:08
zeqHTML5 is the future I hear ;)19:08
freemangordonaah, fennec scores pretty well, 363 or so19:08
freemangordonbetter than my desktop mozilla :P19:09
zeqLOL19:09
zeqfennec/qt?19:09
freemangordonyep19:09
zeqI think there's a page-loading bug with the snapshot we're using, it seems to take too long to start loading pages19:10
zeqthat's pretty impressive, beating your desktop though19:10
freemangordonBTW could you pull the latest source, I saw some memty leaks fixed19:11
freemangordonzeq: well, I have all of the history disabled on my desktop, that is why ;)19:11
zeqI've got a remote setup on my local repo19:11
freemangordonso?19:12
zeqI'm actually test building rebased upstream on upstrean now19:12
freemangordonwould you pull the latest?19:12
*** ArGGu^^ has quit IRC19:12
freemangordonaah, push it please19:12
zeqI'll push it to master if it works :)19:12
freemangordoni saw some gl stuff fixed too19:12
zeqbuild should be finished soon19:13
*** valerius has quit IRC19:13
freemangordonok19:14
*** x29a has quit IRC19:16
freemangordonzeq: damn, there is this toolkit option too "cairo-gtk2-x11"19:16
freemangordoni am lost now19:17
zeqgoodness knows19:17
zeqwhat were you using before?19:17
freemangordoncairo-gtk219:17
zeqyou'd imagine that was the same thing..?19:17
freemangordondon't think so19:18
zeqmaybe in enables x11 extensions outside the scope of gtk2?19:18
freemangordonyep19:19
zeqI'm typing really badly today :/19:19
* SpeedEvil passes zeq a finger-sharpener.19:20
zeqthanks SpeedEvil :)19:20
zeqweird errors though in/it??? I think the problem is higher level than my fingers...19:21
zeqor should that be lower-level? :)19:21
freemangordonyour head or your feet?19:22
*** mvp_ has joined #maemo19:22
zeqseeing if a medicinal beer helps :)19:24
*** Dr_I has joined #maemo19:24
*** Neutron_ has quit IRC19:24
*** Pauly12 has quit IRC19:25
*** x29a has joined #maemo19:27
*** x29a has quit IRC19:30
*** x29a has joined #maemo19:31
*** trbs has joined #maemo19:32
*** muelli has joined #maemo19:32
*** valerius has joined #maemo19:35
*** valerius has quit IRC19:40
*** ALoGeNo has joined #maemo19:48
*** ALoGeNo has joined #maemo19:48
*** paroneayea has quit IRC19:55
DocScrutinizeryummy, medical beer19:57
*** jargon- has quit IRC19:57
*** ArGGu^^ has joined #maemo20:01
*** freemangordon has quit IRC20:01
*** paroneayea has joined #maemo20:01
*** freemangordon has joined #maemo20:03
*** Neutron_ has joined #maemo20:03
zeqfreemangordon: when I said the build should be completed soon, I'd forgotten how long a full build takes!20:05
zeqI think it's finally linking20:06
zeqI do feel much better for a medical beer! ;)20:07
*** APTX has quit IRC20:09
*** APTX has joined #maemo20:12
*** Vlad_on_the_road has joined #maemo20:15
*** saltsa has quit IRC20:17
*** florian has joined #maemo20:21
*** florian has joined #maemo20:21
FIQmicroB's UI isn't really intuitive20:21
FIQI'm not surprised if there's still hidden features that I don't know of, that requires strange cryptic gestures20:21
RiDmicroB had better days20:22
FIQI mean, *after* I''ve learned about the mouse thingy to mark text, it took me several months before I got the hang of doing it correctly20:24
FIQbefore, I just randomly did things, hoping it would work20:24
zeqFIQ: serious question: how would you do it?20:24
FIQI place my finger at the leftmost place of the phone, slowly going right before the mouse activates20:25
FIQthe leftmost, as in, before even the touch screen begins20:25
RiDi do it with the nails20:26
RiDlol20:26
FIQI meant nails :p20:26
FIQmore precision20:26
FIQMost genious gesture ever20:26
FIQWho came up with that?20:26
zeqI mean how would you do it differently?20:27
zeqI actually quite like the UI, but then it did take a while to figure it out, so I can't argue with the intuitiveness of it!20:27
FIQOh, well, I never understanded how you did it, just that it was possible20:27
FIQSo I randomly did things before it did what I wanted20:27
FIQI knew it was something with the left part of the screen, but not more20:28
FIQSo yeah.. :p20:28
zeqFIQ: Imagine, you could start again... What would you do to achieve the same functionality?20:28
*** saltsa has joined #maemo20:28
FIQplace it in the top menu20:28
*** saltsa is now known as saltsa_20:28
FIQit's slower, but at least more available20:29
zeqLike: "Activiate pointer"?20:29
FIQthat, or "mark text" :p20:29
FIQ(as is the widest usage)20:29
zeqit does more that mark text though20:29
FIQYes, it does, but that is the main thing, no?20:29
zeqIt's *a* thing :P20:29
zeqif you need to select certain elements it's often necessary too20:30
RiDor long press, and have a feature for activate pointer20:31
RiDkinda like Opera does20:31
zeqok, that sounds more usable20:31
*** sq-one has quit IRC20:32
zeqI still prefer the MicroB gesture though ;)20:33
RiDMe too, but the long press is not bad either20:33
*** e-yes has quit IRC20:33
FIQthe gesture is definetely usable, it's just.. not obvious at all20:33
FIQgood luck finding the feature on your own, etc20:34
RiDit took me more time to figure out the history browser lol20:35
FIQNot me :p20:35
FIQMostly because I activated it by mistake all the time20:36
*** user has joined #maemo20:36
FIQAnd, tbh, I rarely use back/forward on microB20:36
FIQmostly because it need a page reload20:36
FIQand microB isn't really the fastest browser ever created...20:36
DocScrutinizerhttp://wiki.maemo.org/N900_Shortcuts_and_Gestures20:36
DocScrutinizerFTW20:37
RiDFIQ its starting to get really sluggish on these new pages20:37
*** user is now known as qwazix_90020:37
RiDits pissing me off more and more everyday20:37
FIQMy old phone was a Sony Ericsson P1i20:37
qwazix_900one great problem with microb is20:37
FIQI used Opera Mobile 9.5 on it, and loved it20:37
FIQSure, it's probably slower *today* than microB20:37
DocScrutinizer(exists since... 04:43, 17 January 2010)20:38
RiDyou can use opera on n900 too, but there's no "fullscreen" mode that only gets rid of the top bar20:38
qwazix_900that you cant access history without loading a random page at startup20:38
ShadowJKThere used to be a bookmark in the browser, which took you to a page which explained all those gestures. I think if oyu tap that bookmark today though, you get redirected to advertisements for lumia20:38
FIQbut with that stats, it was awesome20:38
RiDit always hides the bottom bar. which is why i dont use. having to switch fullscreen/windowed all time to change tab is not cool20:38
FIQno refreshing on back/forward, fast, *proper* UI20:38
*** brzys has joined #maemo20:38
RiDqwazix_900 that annoys me too20:38
FIQthe only problem with opera mobile 9.5 on P1i was that it was a beta, and crashed from time to time20:39
qwazix_900I actually like the swipe for pointer but it could use an icon in the menu for new users20:39
FIQand somehow took the OS with it.. a problem I solved when swiss manager elite came out for that OS20:39
FIQafter that, the phone was really decent, and nice, to the level that I sometimes feel that I would prefer it over microB today20:40
ShadowJKEh, you can tap bottom right in opera and it brings up tabs, iirc?20:40
RiDShadowJK and turns off fullscreen.20:40
RiDi once managed to get it fullscreen, without hiding the bottom bar.20:40
RiDthe problem is that i could never reproduce it again20:40
RiDopera scrolling is usually more stutterish, but it always show the page and loads faster on most of the cases20:41
RiDalso it doesn't incredibly bog down when showing large images, unlike microb20:41
ShadowJKWell it goes back to fullscreen once you've used the UI20:41
FIQDelete character. Works in Notes App, entering URLs in Browser. Does not work in Abiword.20:41
FIQwhy was things like this not implemented simply as the key Delete?20:42
*** hurbu_ has joined #maemo20:42
FIQinstead of letting apps implement it on their own20:42
DocScrutinizer??20:42
japhon their own? that's X20:42
RiDShadowJK it makes switching tabs and etc very unpractical for me20:42
ShadowJKoh?20:43
RiDit's one more extra click and a screen "flash"20:44
japhI'm using vimperator in firefox. it's ok20:44
japhfor some reason it didn't want to accept pentadactyl20:44
*** retro|cz has joined #maemo20:45
*** githogori has joined #maemo20:48
*** NIN101 has quit IRC20:55
*** APTX has quit IRC20:56
*** hurbu_ has quit IRC20:58
*** APTX has joined #maemo20:59
*** Hurrian has quit IRC21:01
zeqFIQ: RiD : the limitations of microb are why freemangordon and I are working on Fennec.21:05
FIQI used fennec once, and when it crashed on the exact same page 4 times, I stopped21:06
FIQLagged very much21:06
zeqIdeally we'd like to re-implement a native UI like microb21:06
FIQThat being said, I tried it several months ago, maybe it's better21:06
zeqFIQ: We'll get good performance of out it.21:07
*** Arkenoi has quit IRC21:07
qwazix_900zeq, did you see that new thread on tmo about native ui fennec?21:07
zeqqwazix_900: I'll take a look...21:08
qwazix_900looks like romaxa posted sth about interfacing qt and xul21:08
qwazix_900http://talk.maemo.org/showthread.php?t=8519121:11
*** JakDaRippa has joined #maemo21:13
*** hurbu_ has joined #maemo21:14
RiDzeq well, good luck with that :)21:15
*** accumulator has joined #maemo21:16
*** accumulator has quit IRC21:16
*** Darkchaos has quit IRC21:17
*** eMHa has quit IRC21:18
*** eMHa has joined #maemo21:20
RiD(I was able to tell from my playswoosh sound attached to kbd slide open event)21:20
RiDtell me more about that21:20
RiD(5 hours later)21:21
*** valerius has joined #maemo21:21
zeqfreemangordon: To get html5 video working via gstreamer we need to bump the version.  Can that be done while keeping the binary codecs working?21:22
*** Pali has joined #maemo21:23
*** APTX has quit IRC21:24
zeqfreemangordon: there's a problem with my rebased fennec, it give an error "Object::connect:: No such slot MozQWidget::orientationChanged() ... I'm looking into it.21:24
*** APTX has joined #maemo21:24
freemangordonzeq: push it21:24
freemangordonther is no such slot, because I added it :)21:24
freemangordonin mozqwidget.h21:24
*** paroneayea has quit IRC21:25
freemangordonzeq: the version of what needs to be bumped?21:25
zeqgstreamer21:25
freemangordonno way21:25
*** hurbu_ has quit IRC21:25
freemangordonwhy is that?21:25
zeqnot possible?21:25
freemangordonwhat our gsteramer is missing21:25
*** hurbu_ has joined #maemo21:26
freemangordonsome codec or what?21:26
zeqI haven't tried removing the configure version check to find out21:26
freemangordonaah :)21:26
freemangordonmost probably it will compile21:26
ShadowJKIn general, trying to do h264 in software will have pretty catastrophic performance :P21:27
freemangordonand if we lack some definitions, we could easily add them, i have some experience because of gst-dsp21:27
freemangordonShadowJK: : nod:21:27
zeqfreemangordon: I know it's essential to keep the accelerated codecs working :)21:28
freemangordonzeq: please push the new version :)21:28
freemangordonzeq: we will, i promise21:28
freemangordoneven with our outdated gstreamer21:28
freemangordonwhat we need is gst-dsp and i have some experience with it ;)21:29
freemangordonzeq: in case you are not aware - http://talk.maemo.org/showthread.php?t=77695&highlight=720p21:29
freemangordon:P21:29
zeqJust need a higher res screen ;)21:30
freemangordonno, you don't it is down-sampled21:31
*** klink has quit IRC21:31
zeqI know :)21:31
zeqit's a shame it doesn't have HDMI out though21:31
freemangordonyeah. and it is even bigger shame that it does not have HDMI oou given the HW is capable of21:32
freemangordonare you pushing the new version :P21:32
zeqfreemangordon: can you check the repo is ok?21:32
freemangordondoing git pull21:32
zeqok21:33
freemangordonwidget/qt/faststartupqt/nsFastStartupQt.cpp: needs update21:33
freemangordonwidget/qt/nsWindow.cpp: needs update21:33
freemangordonwidget/xpwidgets/nsBaseWidget.cpp: needs update21:33
zeqgit status21:34
freemangordonhere are your missing functions21:34
zeqdamn21:34
zeqnot sure how that happened21:34
freemangordonI will merge the changes and will put them back21:34
zeqok21:34
zeqthanks21:34
freemangordonzeq, don't know what is wrong on your side, once i reverted changes i did aftel last commit, everything merged just fine21:40
freemangordonthere was some debug code i added21:40
zeqoh ok21:40
freemangordonso there is nothing to push21:40
zequpdate was fine then?21:41
freemangordonyep21:41
zeqthat's something :P21:41
freemangordonyou may try git rest --hard HEAD21:41
freemangordonlleme check if it is still buildable21:41
freemangordongit reset --hard HEAD21:41
*** mardi has quit IRC21:43
freemangordonhmm, don't like that note:21:44
zeqI was previously testing on a local branch.  It could have been resolved when I moved to master21:44
freemangordon# depend on >= 0.10.33 as that's when the playbin2 source-setup signal was21:44
freemangordon:(21:44
freemangordonanyway, it is for the future21:44
freemangordonbuild started, lets see how it will end :)21:45
zeq:fingers crossed:21:46
freemangordonzeq: http://lists.freedesktop.org/archives/gstreamer-commits/2011-February/047159.html21:46
freemangordonwe can just backport that (if it is not already there21:47
freemangordon)21:47
freemangordonalso i like this note from the patch: This is functionally equivalent to connecting to the notify::source signal, but more convenient.21:47
freemangordonit tells me that we ca workaround it for maemo521:48
freemangordons/ca /can /21:48
infobotfreemangordon meant: it tells me that we can workaround it for maemo521:48
*** klink has joined #maemo21:49
zeqit doesn't seem insumountable21:49
*** Pali has quit IRC21:50
freemangordonwhat a compilicated word, thank you google translate :D:D:D21:50
zeq:)21:50
*** Pali has joined #maemo21:50
freemangordonzeq: btw which fennec version is that?21:54
freemangordon16a1 again?21:54
zeqI don't think they bumped it21:54
freemangordonok21:55
freemangordonPali: any progress?21:56
*** Pali has quit IRC21:56
freemangordon:(21:56
*** Mrick0363 has quit IRC21:56
freemangordonWTF?21:56
freemangordonzeq: saw my query?21:57
zeqfreemangordon: same version AFAIK21:57
freemangordonfor fennec?21:58
zeqseems to be21:58
freemangordonaah, ok21:58
freemangordonagain: did you see my query (query as PM in IRC)21:58
*** mardi has joined #maemo21:59
*** retro|cz has quit IRC22:01
*** NIN101 has joined #maemo22:03
*** Pali has joined #maemo22:08
*** Timmy has joined #maemo22:09
*** Arkenoi has joined #maemo22:09
Timmywhen i try to flash a kernel via x-term i face this problem: could not connect to '127.0.0.1' what's the problem?22:10
*** rm_you has quit IRC22:10
*** valerius has quit IRC22:17
*** robink has quit IRC22:18
*** robink_ has joined #maemo22:22
*** Timmy has quit IRC22:25
*** bef0rd has joined #maemo22:25
*** dhbiker has quit IRC22:33
*** robink_ is now known as robink22:34
*** robink has quit IRC22:34
*** robink has joined #maemo22:34
*** Timmy has joined #maemo22:37
*** Timmy has joined #maemo22:37
*** dhbiker has joined #maemo22:37
*** nox- has joined #maemo22:38
*** valerius has joined #maemo22:39
*** M4rtinK2 has joined #maemo22:43
*** M4rtinK has quit IRC22:43
*** k273 has joined #maemo22:45
*** buntfalke has joined #maemo22:48
*** Arkenoi has quit IRC22:50
*** dos1 has quit IRC22:51
*** dos1 has joined #maemo22:52
*** dos1 has quit IRC22:54
*** valeriusN has joined #maemo22:56
*** paroneayea has joined #maemo22:58
freemangordonzeq: this version is broken :(22:59
freemangordonwithouw HW acceleration it is slow23:00
freemangordonneed to find what broke it23:00
zeq:(23:02
zeqbisection is going to be fun23:03
freemangordonyeah23:03
* SpeedEvil wonders if chronological bisection would be better.23:04
SpeedEvilAs most of the time, your subsystem is not going to be the one being patched23:04
zeqSpeedEvil: the trouble is the build-time ...23:04
SpeedEvilI guess.23:04
zeqSpeedEvil: I mean whatever method23:04
*** perlite has quit IRC23:05
zeqwe need to look for a suspicious commit23:05
*** Timmy has quit IRC23:05
*** nmjnb has quit IRC23:05
*** perlite has joined #maemo23:06
SpeedEvilIs this kernel?23:07
freemangordonfennec23:08
SpeedEvilah23:08
SpeedEvilOh23:08
freemangordoneven worse :)23:08
SpeedEvilthat does take a rather longer time to build.23:08
SpeedEvilAssuming it's about the same as firefox23:08
zeqespecially in scratchbox!23:08
SpeedEvilI assume you can't duplicate the bug in x86, running remotely?23:09
freemangordonyeah :(23:09
freemangordonSpeedEvil: nfc23:09
freemangordonzeq: the good news is that HW acceleration is smoother and crashes rarely23:10
freemangordon(i.e. when enabled)23:10
zeqwith qt?23:10
freemangordonyep23:10
zeqok, that's something :)23:10
*** qwazix_900 has quit IRC23:10
zeqobviously being worked on23:10
zeqmaybe that's what broke non-hw accel23:10
zeq:S23:11
freemangordoncould be23:11
*** qwazix_900 has joined #maemo23:11
freemangordonanyway, I am afk for a while23:11
zeqfreemangordon: np23:11
freemangordonzeq; I wonder should we prepare a patch agains previous tree23:12
freemangordon(and who will prepare it :) )23:12
* freemangordon is afk23:13
*** Arkenoi has joined #maemo23:14
*** _federico3 has quit IRC23:16
*** Pali has quit IRC23:20
*** ced117 has quit IRC23:24
*** ArGGu^^ has quit IRC23:29
*** zeq has quit IRC23:32
*** dos1 has joined #maemo23:34
*** mardi has quit IRC23:38
*** e-yes has joined #maemo23:46
*** steve__ has joined #maemo23:47
*** saltsa_ has left #maemo23:47
*** Vlad_on_the_road has quit IRC23:47
*** mardi has joined #maemo23:51
*** qwazix_900 has quit IRC23:57

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!